Chronic graft-versus-host disease (GVHD) is one of the most devastating long-term complications after infusion of allogeneic hematopoietic stem cells, and it remains one of the major barriers to successful transplantation. Relatively little progress has been made in understanding and improving treatments for chronic GVHD over the last 20 years, and the survival rate after diagnosis of chronic GVHD has barely improved despite advances in supportive care. The NIH convened a Consensus Conference on this topic in June 2004 and recently published its recommendations on improving research methods in a series of six papers. The Consensus Conference recognized a lack of basic observational data to guide standards for clinical trials. In order to address this need, we propose to establish a multi-center, observational, longitudinal cohort in order to improve outcomes assessment in chronic GVHD with the specific aims of (1) validating prognostic factors for risk stratification;and (2) defining significant variables for a chronic GVHD activity index that predicts short-term (provider perception of change, patient perception of change, and changes in immunosuppressive medications) and longer-term outcomes (overall survival, time to discontinuation of systemic immunosuppressive therapy, and functional impairments). This goal will be accomplished by assembling a large modern cohort of people with chronic GVHD at four large transplant centers. Approximately 880 people (half prevalent cases, half incident cases) with chronic GVHD will be enrolled. Every 3-6 months we will collect both objective and subjective measures reflecting disease activity, response to therapy, detailed physician assessments about organ involvement, and patient self-assessments about symptoms, functional status, and quality of life. Data will be used to test published hypotheses and the new recommendations emanating from the NIH Consensus conference. We will also be able to provide the detailed data needed to understand modern trends in chronic GVHD incidence, manifestations, and response to treatment. These studies are needed to operationalize the recommendations of the NIH Consensus conference, advance our understanding of chronic GVHD, and enhance our ability to conduct clinical trials.